The goal of this study is to test the effectiveness of a 10 week preventive intervention for bereaved parents following the sudden accidental deaths of their adolescent and young adult children. The study rationale is derived from the logic of preventive intervention and conflicting propositions of stress/support sequencing which suggests that stressful concerns linked with changes resulting from bereavement loss may be temporally ordered and therefore require different dimensions of support at different times. The specific aims are to: 1) determine the immediate post-intervention and short-term effects of a two-dimensional supportive intervention provided to bereaved mothers 2 to 7 months post-loss; 2) Explore the immediate and short-term effects of the two-dimensional preventive intervention upon three groups of fathers: those who participate in the intervention; those whose wives participate in the intervention, and those in which neither member of the dyad participate; 3) Identify factors associated with less impaired role functioning fewer disruptions in health maintenance behaviors, less mental distress and progressive loss resolution among bereaved mothers and fathers, and 4) Assess the perceived value of information-focused and emotion-focused support by participant responses and group leader clinician and independent observer ratings. A randomized experimental design with pretest and repeated posttest observations will be used in which 200 bereaved parents will be enrolled in 12 week cycles in groups of 6 to 12 over a two year period and then followed 6 additional months. Parent dyad participant will be randomly assigned to intervention (n=100) and control (n=100) conditions. Baseline data will be collected, the treatment administered, and post-intervention data will be collected immediately and at 6 months post-intervention. Mothers' scores will be used to test the study hypothesis. The results are expected to suggest whether one dimension of support, i.e. information-focused or emotion-focused, is perceived by parents as more helpful than the other; whether the intervention affects role performance, health maintenance behaviors, mental distress, and loss resolution and whether effects are sustained or diminished over time. The effect of three types of father participation on outcome variables will be assessed. The unique aspects of the study are that marital, parental, and occupational role performance and health maintenance behaviors are dependent variables in addition to measures of health status, emotional distress and grief resolution.